Toward Skynet: Johns Hopkins University Improves War Simulations with Artificial Intelligence
Johns Hopkins University in Maryland is preparing to upgrade its AI-powered wargaming tools to help the Pentagon identify adversaries’ weaknesses in real-world conflicts. The work, conducted at the university’s Applied Physics Laboratory (APL), involves upgrading two systems, Generative Wargaming (GenWar) and the Strategic AI Gaming Engine (SAGE), using proprietary data for Department of Defense programs.
Wargaming, used to practice decision-making in complex and uncertain environments, remains a key tool for analyzing human behavior in complex and uncertain contexts through experiential learning. However, traditional wargaming requires skilled facilitators and complex design, which reduces the speed and scalability of the process.
GenWar combines generative AI, modeling, simulation, and human expertise to create and execute scenarios in days instead of months, analyze dozens of alternative futures, and focus human experts on the most meaningful scenarios.
According to Andrew Mara, director of the National Security Analysis Office at the APL, the Pentagon has been looking for solutions like this for over a decade. Now, he believes, necessity and technology have met, and the combination of cutting-edge technology and an experienced team could change the very nature of war gaming.